Amazon tells staff to stop using AI just to use AI after shutting down its token leaderboard

The tech giant has retired its employee-generated “KiroRank” ranking system, concluding that it was forcing staff to increase AI activities rather than solving real business problems, as major companies in Silicon Valley began to reassess the true value of out-of-control AI spending.

Amazon has shut down its employee-created AI leaderboard that tracks token consumption across its workforce after a senior executive issued a direct warning to staff not to use AI technology just to improve their position on the dashboard rather than solving real business challenges.

The internal ranking tool, known as “KiroRank,” was created unofficially by a group of Amazon employees and tracked AI token usage as an indicator of productivity. The system was discontinued amid concerns that it encouraged a behavior now commonly known in Silicon Valley as “tokenmaxxing”: the practice of measuring AI-driven output by pure consumption volume rather than any meaningful outcome.

What Amazon’s Senior Vice President Says About the Misuse of AI at the Company

The company’s displeasure with the leaderboard became public after Dave Treadwell, Amazon’s senior vice president, directly addressed employees. “Please don’t use AI just for the sake of using AI,” Treadwell told staff earlier this week. “Use AI to help you solve customer problems, solve business problems, and innovate.”

An Amazon spokesperson confirmed Business Content Stating that KiroRank had been “deprecated”, he explained that the dashboard was an unofficial initiative and “was never intended to promote the use of AI for the sake of use”.

Amazon Says Teams Retain Freedom to Pursue AI Use in Their Own Way

Although Amazon has shut down KiroRank, it hasn’t completely banned AI productivity tracking. Individual teams across the organization retain the freedom to determine how they will deploy AI tools and measure that usage internally.

What the company clearly steers clear of is the idea that a higher token count indicates greater efficiency or innovation. Amazon tracks AI token consumption for cost management purposes, but the spokesperson made it clear that the company does not promote tokenmaxxing as a philosophy.

“We are focused on embracing AI and sharing best practices to celebrate innovation and operational efficiency gains across the company, and we are proud of how our teams are embracing this technology,” the spokesperson said.

In April, it was reported that Amazon’s retail division was closely monitoring how engineers across the business were interacting with AI tools, including monthly adoption rates, how deeply AI capabilities were embedded into workflows, and what tangible results those deployments produced. In other words, the focus was on results rather than raw activities.

Big Tech Is Rethinking AI Spending As It Can’t Keep Up With Returns

Amazon’s move comes at a time when a broader recalibration is taking place in the tech industry. After years of largely uncontrolled enthusiasm for AI investment, some of the world’s most prominent companies are beginning to question whether financial spending is producing commensurate productivity gains.

Andrew Macdonald, Uber’s chief operating officer, recently acknowledged that the ride-sharing giant has failed to see productivity and other gains in line with its increased AI spending. His comments came after a moment that drew quite a bit of attention in tech circles: Praveen Neppalli Naga, Uber’s chief technology officer, announced that the company had already exceeded its entire annual Claude Code budget by April.

All these revelations suggest that the era of liberal AI investment justified by expected future returns is beginning to lead to stricter scrutiny of what spending actually produces.

What is Tokenmaxxing and Why Is It Falling Out of Favor?

Tokens are the basic units through which large language models process and produce language, and serve as the building blocks of AI chatbots, coding assistants, and automated agents. Token usage in the industry has increased sharply through 2026, driven by the proliferation of agency AI systems that can operate autonomously with minimal human supervision for significantly longer periods of time.

As this usage has grown, so has the tendency to treat token consumption as a performance metric, a number that is easy to track, report upstream, and optimize regardless of whether the underlying work it represents is actually useful.

The KiroRank section shows exactly this risk. When employees are ranked according to how much AI they use, the incentive structure rewards use, not value. Amazon’s decision to shut down its leaderboard and make it public is a sign that at least one of the world’s largest tech companies is no longer comfortable with that tradeoff.